# What is context in AI?

Official page: https://aitje.com/en/knowledge/wat-is-context

> Context is all the additional information you provide to a model so that it better understands what you mean. Good context makes the difference between a generic answer and something that really suits your task, organisation or document.

### What constitutes context?

Context may consist of the question itself, previous messages, documents, instructions, examples, roles, company information or desired output. In practice, anything that helps the model provide a more targeted response forms part of the context.

### Why is context so important?

A model without context usually provides a general answer. As soon as you add your objective, source information and constraints, the same AI becomes much more specific. The quality of an answer therefore often depends less on the model alone and more on the context you provide.

### What types of context are there?

You might consider task context, such as exactly what the user wants; document context, such as relevant sources; and organisational context, such as tone of voice, policy or process agreements. The better these layers align, the more useful the output becomes.

### What often goes wrong?

Many applications provide too little, too much or the wrong context. Too little context results in superficial answers. Too much context makes the answer slow or muddled. The wrong context causes a model to go in the wrong direction, even if the question itself is logical.

### A practical rule of thumb

Provide only what is necessary to carry out the task properly. Think in terms of the decision or answer you wish to receive, rather than the urge to dump all the information at once. Good context is focused, relevant and up to date.

## Contact

[Contact AITJE](https://aitje.com/en/contact) — contact@aitje.com